Human respiratory syncytial virus (RSV), an enveloped RNA- containing virus of the paramyxovirus family, is the most important viral agent of pediatric respiratory tract disease. Its genome is a single negative strand of RNA of 15,222 nucleotides that encodes ten major mRNAs. The viral genes are transcribed in a sequential fashion in which transcript abundance decreases with increasing distance of its gene from the promoter (polar transcription). The purpose of this project is to identify the functions of the viral proteins and to reconstruct events in the viral growth cycle under conditions where they can be more readily studied. Knowledge of the functions of the proteins is important for the design of live- attenuated recombinant vaccine viruses. We previously developed an intracellular transcription and replication system for RSV based on components expressed from transfected plasmid-borne cDNAs (see accompanying report). This involves transfecting tissue culture cells with plasmids which individually encode a helper-dependent minireplicon analog of negative-sense genomic or positive-sense antigenomic RNA as well as whatever mix of RSV proteins is desired. Three proteins, the nucleocapsid N protein, phosphoprotein P and large polymerase subunit L protein, constitute the RSV replicase whereas, unexpectedly, the transcriptase requires in addition expression of the M2 mRNA. The M2 mRNA contains two overlapping translational open reading frames (ORFs), and the upstream one (M2-1) was shown to encode a processivity factor. This is a completely novel finding for the negative strand RNA viruses, and among RNA viruses such a factor has only been described for human immunodeficiency virus 1, namely the Tat protein. Additional experiments indicated that the presence of M2-1 is not required for correct initiation, termination or polyadenylation. However, in addition to its processivity activity, the M2-1 protein was found to reduce the efficiency of transcriptional termnation at the gene-end (GE) signal and thus increases the amount of transcriptional readthrough. The effect of M2-1 on RSV sequential transcription is being examined using minigenomes which contain authentic groups of viral genes. Both activities, namely processivity within genes and partial readthrough of GE signals, has the effect of reducing the gradient of transciptional polarity. Transcription and the first step of RNA replication each involve the synthesis of positive-sense RNA off the genomic template. It is generally assumed that the two processes are interrelated and are regulated by the availability of N protein to direct encapsidation of the antigenome. In the absence of sufficient N protein, transcription would predominate. An alternative model is that the M2-1 transcription factor might have a regulatory effect by favoring transcription over replication. The reconstituted transcription/replication system provided the first opportunity to directly test these models. A series of tightly controlled experiments showed that the levels of transcription and RNA replication are insensitive to changes in the levels of N or of the other nucleocapsid-associated proteins. The available data suggests that these processes are independent rather than tightly interrelated. Expression of the second ORF of the M2 mRNA (M2-2) inhibits RNA replication and transcription, providing functional evidence that this represents an eleventh RSV gene and is a novel negative regulatory factor. We presently are working to detect this protein in infected cells and to determine the mechanism by which it is expressed from its overlapped ORF. The nonstructural protein NS1 also was found to be a potent inhibitor of RSV RNA synthesis. Like the M2-2 protein, NS1 strongly inhibited transcription and both steps of RNA replication. We found that it is possible to reconstitute virion morphogenesis by coexpression of appropriate envelope components with the above-mentioned transcription and replication system. Morphogenesis was assayed by the passage of a minireplicon to fresh cells. This showed that the matrix protein M, attachment glycoprotein G and fusion glycoprotein F protein are important for the formation of transmissible virus-like particles. The small hydrophobic SH protein, both ORFs of the M2 mRNA, and the nonstructural NS1 and NS2 proteins are completely dispensable for particle formation. Interestingly, coexpression of the M2-1 transcription factor reduced the efficiency of virion packaging, suggesting that a nucleocapsid which is engaged in fully processive transcription might be refractory to packaging. Consistent with this, the further coexpression of the M2-2 inhibitory factor increased the efficiency of packaging and countered the effects of M2-1. Thus, M2-2 might be a packaging factor.